Year: 2025
Type: article
Abstract: Blockchain networks face significant scalability challenges, particularly in handling micropayments efficiently. Current Layer-2 solutions, while promising, suffer from high channel opening costs and suboptimal routing efficiency, limiting their practical adoption for micropayment scenarios. This paper presents a novel Layer-2 scaling solution that leverages graph neural networks (GNNs) for optimal payment routing combined with predictive channel balancing mechanisms. Our approach ad- dresses the critical research gap in optimization algorithms for state channel routing and liquidity management. We introduce a comprehensive framework that integrates advanced machine learning techniques with blockchain technology to achieve substantial improvements in transaction throughput and cost reduction. Through extensive simulation and real-world testing, our solution demonstrates a 10-fold reduction in micropayment costs while maintaining instant settlement capabilities. The pro- posed system achieves an average transaction confirmation time of 0.2 seconds with fees reduced to 0.01% of transaction value. Our contribution includes a novel GNN-based routing algorithm, a predictive channel rebalancing mechanism, and a comprehensive evaluation framework that validates the effectiveness of our approach across various network topologies and transaction patterns.
Institution: Manipur International University
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4414537874